The mismatch repair pathway reduces genome instability and oncogenesis by correcting replication errors. However, mismatch repair may not function at all genomic sequences with equal efficiency. Our goal is to determine why mismatch repair fails to respond to damage at certain sites in the genome, explaining why some loci are hot spots for mutation and rearrangement. In particular, repetitive guanine-rich sequences are found at unstable proto-oncogenes and recombinogenic loci. Guanine repeats have the unusual ability to adopt four-stranded structures in vivo and in vitro, called guanine quadruplex or G4 DNA. G4 has the highest potential to form when repetitive guanine-rich DNA is transiently liberated from cytosine-rich complement during transcription or replication. It is unclear why G4 structures are prone to instability, but it is clear that several DNA maintenance activities in the cell respond to structure formation. Based on observations that loci with repetitive guanines are unstable, we asked if DNA repair functions normally within guanine-rich DNA. We used a well-established assay for human mismatch repair, which relies on human nuclear extracts and defined substrate molecules. We discovered that GT mismatched DNA is not efficiently corrected by mismatch repair in some repetitive sequence contexts. Specifically, correction is poor when repetitive guanine sequences serve as the template for repair synthesis. These results indicate that repetitive guanine motifs interfere with human mismatch repair, leading to genome instability. We plan to determine why, and we will accomplish our goal with two specific aims. In Aim 1, we will determine the stage of human mismatch repair that is blocked by repetitive guanine by characterizing the repair templates that are inhibitory. We will also directl test the model that mismatch repair is blocked because of stalled synthesis, resulting in single-stranded intermediates. Aim 2 will determine if the block to repair is due to G4 structures or to guanine repeats. We will also assay other repetitive loci to define genomic sequences and repeat motifs that are inherently unstable and refractory to high fidelity mismatch repair. The ability of alternative DNA structures to inhibit DNA repair has major implications for clarifying mechanisms of genome instability and oncogenesis. Our studies will provide a new understanding of genome instability mechanisms, connecting for the first time human mismatch repair responses with repetitive sequences and non-duplex DNA structures.